In virtual tests, Ewers’ algorithm outperformed both approaches on two key measures. The distance a drone must fly to find a missing person, and the percentage of time the person is found. While the lawnmower and traditional algorithmic approaches found people 8% and 12% of the time, respectively, Ewers’ approach found people 19% of the time. If proven successful in real-life rescue situations, the new system could accelerate response times and save more lives in scenarios where every second counts.
“The search and rescue area in Scotland is very diverse and very dangerous,” says Ewers. Emergency situations can arise in the dense forests of the Isle of Arran, the steep mountains and slopes around the Cairngorm Plateau and the surface of Ben Nevis, one of Scotland’s most revered but dangerous rock climbing destinations. there is. “If we can send a drone to search efficiently, it could potentially save lives.”
Search and rescue experts say that using deep learning to design more efficient drone routes could help find missing people faster in various wilderness areas, depending on how well the environment lends itself to drone exploration (drones may be more likely to choose dense areas rather than open ones). Navigating the canopy is more difficult). For example brush).
David Kovar, director of the National Association for Search and Rescue, says: “This approach in the Scottish Highlands certainly appears viable, especially in the early stages of a search when you are waiting for someone else to turn up.” . Williamsburg, Virginia, which has used drones for everything from disaster response in California to wilderness search missions in the White Mountains of New Hampshire.
However, there is a caveat. The success of these planning algorithms depends on how accurate the probability maps are. Overreliance on these maps can result in drone operators spending too much time searching the wrong area.
